PAC Confidence Sets for Deep Neural Networks via Calibrated PredictionDownload PDF

25 Sept 2019, 19:21 (modified: 11 Mar 2020, 07:34)ICLR 2020 Conference Blind SubmissionReaders: Everyone
Original Pdf: pdf
Abstract: We propose an algorithm combining calibrated prediction and generalization bounds from learning theory to construct confidence sets for deep neural networks with PAC guarantees---i.e., the confidence set for a given input contains the true label with high probability. We demonstrate how our approach can be used to construct PAC confidence sets on ResNet for ImageNet, a visual object tracking model, and a dynamics model for the half-cheetah reinforcement learning problem.
Keywords: PAC, confidence sets, classification, regression, reinforcement learning
10 Replies